looking for "optimal weighting" algorithm

Duncan Smith buzzard at urubu.freeserve.co.uk
Thu Apr 10 20:19:12 CEST 2003

"Terry Reedy" <tjreedy at udel.edu> wrote in message
news:HzGdnQNo1Z2-AAijXTWcog at comcast.com...
> "Stephan Diehl"
> > You are probably looking for "linear regression".
> Linear regression is designed for continuous, not binary outcome
> variables.
> "Duncan Smith"
> > But linear discriminant analysis and logistic regression would give
> you
> Logistic regression use 0-1 outcomes to fit a very particular model: p
> = exp(s)/(1+exp(s)) where p is a probability of 1 versus 0 and s is a
> linear sum of 'independent' variables.  It is typically used for
> modelling probability of response within one group.  It is less likely
> to be useful for two-group classification.
> Terry

Yes, you do have the link function as well as the linear predictor, and the
cut-off would be 0.5 rather than 1.  But in terms of absolute efficiency
there is (according to the literature) little to choose between linear
discriminant analysis and logistic regression.  Changing the cut-off would
be a simple way of attempting to minimise Alex's 'error-cost'.


More information about the Python-list mailing list